[Psychophysiological basis of smells].

There has been an increase in the use of psychophysiological techniques, especially event-related potentials (ERPs) to evaluate the effects of odor on the central nervous system. In the study of ERPs related to odor, there are two main methods of the odor presentation: (1) to investigate the olfactory ERP (OERP) or chemosensory ERP (CSERP), that is to calculate electrophysiological responses to odor stimulation as a trigger; (2) to compare ERP of mental functioning during the odor exposed condition with that of no odor exposure. The amplitude of contingent negative variation (CNV) varies with the odorant being present, when the odor is presented as a trigger and when it is administrated during a task. It is suggested that CNV changes depend on not only the odorant but the anticipation, expectation and emotion of the subject. The latency of the N1 component becomes shorter with increasing concentrations of odors. The N1 amplitude does not depend on the odor concentration while the amplitude of the auditory N1 strongly depends on the stimulus intensity. There is only one report that the olfactory mismatch negativity (MMN) is present in the CSERP. On the other hand, auditory MMN is not affected by odor administration. In relation to the P3 component, rare odors evoked a larger amplitude in contrast to frequent odors, as well as P3 evoked by other modalities. These ERP studies with odors are expected to be applied to such clinical settings as the differentiation between the anosmic patients and normosmic persons, the functional evaluation of patients with brain tumors, the earlier detection of dementia, and the objective evaluation of aromatherapy.

Transient gamma-band response is dissociated from sensory memory as reflected by MMN.

The auditory gamma-band transient oscillatory response has been considered to reflect early cognitive processing and attention triggering, as has been suggested of the mismatch negativity (MMN). We examined whether the auditory gamma-band response was related to sensory memory as reflected by MMN. During the electroencephalogram (EEG) recordings, approximately 2000 click sounds were presented to nine healthy adult subjects with constant SOA of 120 or 170 ms in an ignored condition. At a probability of 10%, a click sound was randomly omitted from the stimulus sequence. EEG epochs responding to omitted clicks and to click sounds were averaged for analysis, respectively, and then those were convoluted by Gabor wavelet for the gamma-band response calculation. The MMN to a deviant omission in a sequence of click sounds was elicited with SOA of 120 ms which was shorter than the duration of temporal window of integration, whereas no MMN was elicited with SOA of 170 ms. In contrast with the MMN, the transient gamma-band response clearly commenced after the stimuli but not after the omissions, regardless whether SOA was short or long. The findings indicate that the brain process underlying the transient gamma-band response should be dissociated from the sensory memory function.

Time may be compressed in sound representation as replicated in sensory memory.

The flow of time in the real world may differ from the one in the neural representation of auditory scene stored in the sensory memory, because the encoded information is free from the rule of actual time. The sensory memory underlying the automatic discriminative system is reflected by the mismatch negativity (MMN). The time-wise image of preceding sounds is integrated into the sensory memory as a 160-170 ms epoch. We measured the mismatch field (MMNm) and the reaction time responding to the omitted segments incorporated into a complex sound. The main result was that the encoded period from early to late omitted segment was shorter than the actual one, suggesting that time may be compressed in the sound representation.

Sound perception affected by nonlinear variation of accuracy in memory trace.

The mismatch negativity (MMN) reflects the neural representation of the acoustic environment stored in sensory memory. The short phase of sensory memory corresponds to the temporal window of integration (TWI) which integrates the neighboring sounds into a unitary event. We measured the magnetic MMN (MMNm) and the discriminative reaction time (RT) responding to an omitted segment incorporated into a complex sound. Consequently, for the late omitted segments, the MMNm amplitudes were decreased, and those latencies and the RTs were prolonged. The percentages of the correct responses were also reduced for the late omitted segments. In sum, the discriminative sensitivity nonlinearly declined toward the end of the TWI, suggesting that the time-wise accuracy of the neural trace nonlinearly varies in sensory memory.

Attentional distraction of CNV depending on the spatial focus.

To investigate the distraction of spatial attention to the task-irrelevant visual stimuli, contingent negative variation (CNV) was measured by using a forewarned reaction time task in 20 healthy subjects. The lasting emission of light, irrelevant to the CNV task, at each perimetric angle of 15 degrees, 30 degrees or 45 degrees to the fixated point was presented to the subjects. The amplitude of early CNV was small only under the light-emission at the angle of 30 degrees. Our results indicate that attention is distracted even by the lasting, task-irrelevant stimuli and that distraction is dependent on the focusing function of attention.

Spectrotemporal window of integration of auditory information in the human brain.

The human auditory system is adapted to integrate temporally successive sounds into meaningful entities, that is, acoustic information units. Hence, sound sequences falling within the temporal window of integration should be coded holistically as unitary representations in the human auditory cortex. Although it is well established that the auditory system operates in the frequency-temporal domain, many previous studies only focused on the temporal domain of the window of integration. Therefore, in the current study we investigated the relationship between the short-term temporal integration and the frequency integration. Event-related magnetic fields in response to infrequent omission of the second tone in repetitive tone pairs composed of two closely spaced tones of different frequencies were recorded. This omission elicited the magnetic counterpart (MMNm) of the electric mismatch negativity (MMN), a change-specific component mainly generated in the auditory cortex, when the interval between the two successive tones was extremely short or when the frequency difference between the two tones was small. These findings suggest that two stimuli presented in close succession might be represented in the auditory system as a unitary integrated event. In addition, as the distance between the two successive tones decreased in the spectrotemporal dimensions, the magnitude of the MMNm increased. Behavioral data also supported these neurophysiological phenomena. This work shows the first neurophysiological evidence that the two-dimensional (spectrotemporal) window of integration, which provides important constraints for the neural processing of the acoustic environment, exists in the human brain.

Impairment in activation of a frontal attention-switch mechanism in schizophrenic patients.

The present study addresses the difference in activities of frontal and temporal mismatch negativity (MMN) generators between healthy controls and schizophrenic patients. Auditory MMNs were measured from 13 medicated schizophrenic patients in a post-acute phase and 12 healthy controls. The probabilities of the standard stimuli were, in different experimental blocks, 95, 90, 80 or 70%. The mean amplitude of the MMN recorded at Fz was significantly smaller in schizophrenic patients than healthy controls only in the conditions with high probability of standard stimuli, while that recorded at mastoid sites was not different in any condition. The present study suggested that schizophrenic patients might fail to cause involuntary attention switch to stimulus change reflected in the lowered MMN amplitude recorded at Fz; whereas the patients might index an adequate detection of the deviant event reflected by the similar amplitude of MMN recorded at mastoid sites.

Differential effects of the hiba odor on CNV and MMN.

The aim of this study was to objectively measure some psychophysiological effects of odors, particularly the effects of odors on the contingent negative variation (CNV) and the mismatch negativity (MMN). It is generally believed that CNV reflects arousal processes and MMN reflects activity in an automatic detection system. Sixteen females were exposed to the odor of Thujopsis dolabrata (hiba), a conifer. CNV was obtained with a foreperiod of 2 s in a traditional click-flash reaction-time (RT) paradigm. Auditory MMN was measured while the subject was reading a book with SOA fixed at 500 ms. The amplitude of the early and late CNV components were significantly larger and RT to the imperative stimulus (IS) was shorter in the aroma condition than in the absence-of-odor condition, but there was no significant difference in the amplitude of MMN obtained in the two conditions. Our results indicate that the odor generates a high level of arousal within the nervous system but does not have a significant effect on automatic information processing.

Early contingent negative variation (CNV) shows a small symmetrical negativity in a somatosensory paradigm.

The influence of sensory modulation on the early stage of information processing was investigated with a somatosensory contingent negative variation (CNV) paradigm. Whether or not even a somatosensory input as well as auditory or visual stimulus to one hemisphere elicits the symmetrical "early CNV" was also examined. Eleven normal individuals (3 males, 8 females) performed a conventional CNV paradigm with a click sound as the warning stimulus (WS) and a red light flash as the imperative stimulus (IS). Nine individuals (5 males, 4 females) did the somatosensory CNV paradigm with paired electrical stimuli as WS and IS. The subjects were instructed to press a button in response to IS as fast as possible. The early CNV amplitude was smaller and P300 latency was longer in somatosensory paradigm than conventional paradigm. In addition, the latency of P100 in a somatosensory paradigm was longer than that of N100 in a conventional paradigm. These findings suggest that the initiation of early detection, reflected by P100, the initiation of cognition, reflected by P300, and orienting response, reflected by early CNV, are delayed in a somatosensory CNV paradigm. Furthermore, all event-related potentials (ERPs) evoked by somatosensory stimuli showed a bilateral symmetry.

The difference in Mismatch negativity between the acute and post-acute phase of schizophrenia.

In order to investigate the trait and state aspects of Mismatch negativity (MMN) amplitude reduction in schizophrenia, auditory MMNs were measured from 13 schizophrenic patients on two occasions, initially when they showed acute exacerbation and later when their symptoms improved. Patients exhibited reduced mean amplitude of the MMN recorded at Fz. There were no significant changes in the amplitude of MMN at Fz between the acute patients and the post-acute patients, despite significant improvement in symptomatology. However, the acute patients showed a significant attenuation of MMN recorded at both mastoids as compared with the post-acute patients. Although the findings of the MMN at Fz support the overall longitudinal stability of MMN deficits in schizophrenia, the acute phase patients showed a modestly altered MMN activity compared with the post-acute phase patients, suggesting that there is some state-dependent modulation of these deficits.